Modeling heart rhythm using human engineered heart tissues

Abstract

Heart rate is both an indicator and modulator of cardiovascular health. Prolonged elevation in heart rate or irregular heart rhythm can trigger the onset of cardiac dysfunction, a condition termed ‘tachycardia-induced cardiomyopathy’. While large animals have historically served as the primary model for studying this condition owing to their similar resting heart rates to humans, their use is limited by cost and throughput constraints. We recently developed the first engineered model of tachycardia-induced cardiomyopathy to overcome this technical bottleneck. Our model uses matured human engineered myocardium coupled with programmable electrical stimulation to emulate the pathophysiological changes in human heart rhythm. This in vitro model, capable of acutely and chronically modulating both beating rate and rhythm, recapitulated the clinical hallmarks of tachycardia-induced cardiomyopathy, and its utility was further validated via molecular comparisons against data from a canine model and human patients. Moreover, this model has improved the throughput and relevance to human genetics, enabling deep mechanistic explorations that were previously impossible. Here we present a comprehensive workflow detailing the fabrication and maturation of human engineered heart tissue, assembly of the electrical pacing system, functional analysis using open-source software and preparation for proteomic and transcriptomic analyses. This 5-week Protocol could be implemented by an experienced bench scientist with strong expertise in cell culture, ideally involving stem cell-derived cardiomyocytes. Given the broad implications of heart rhythm alterations in various cardiac conditions, this workflow can be employed with other biophysical and chemical cues to generate more complex and physiologically relevant cardiac models.

Introduction

Human heart rhythm is an essential modulator of cardiac health. A persistent elevation in resting heart rate in individuals without preexisting heart disease is an independent risk factor for heart failure^1,2. A prolonged fast heart rate of over 100 beats per minute (bpm) alone can induce rapid deterioration of cardiac contractile function, known as tachycardia-induced cardiomyopathy³. Conversely, lowering resting heart rate with selective medications, such as ivabradine, has been shown to improve cardiovascular outcomes in patients with heart failure, and the extent of this improvement is correlated with the degree of heart rate reduction achieved⁴. Independent of heart rate, the regularity of the heart rhythm also plays a key role in cardiac physiology. For instance, cardiomyocytes (CMs) subjected to irregular pacing show impaired calcium handling and increased fibrotic signaling compared with cells paced at a regular rhythm^5,6.

Heart rhythm also plays a critical role in physiological adaptation. Exercise training acutely increases the heart’s workload, mainly by elevating the heart rate by two- to four-fold. This heart rate elevation is driven by a combination of parasympathetic withdrawal and increased adrenergic tone, such as norepinephrine, which enhances cardiac output to meet the metabolic demands of exercise^7,8. Unlike sustained tachycardia or arrhythmia, which are associated with cardiomyopathy, heart rate elevation during exercise is often associated with physiological hypertrophy and improved cardiac function⁹. This suggests that properties such as the duration, intensity and regularity of heart rhythm, combined with the dynamics of adrenergic tone, may underlie the divergent outcomes between pathological and physiological cardiac hypertrophy.

Large animal models such as dogs have been the primary option for studying the role of heart rhythm in cardiac pathophysiology, as they have resting heart rates similar to those of humans. In 1969, dogs were used to model atrial flutter via coronary sinus pacing¹⁰. In 1970, a congestive heart failure model was created in dogs by rapid electrical pacing, and it remains the standard model of tachycardia-induced heart failure today^11–13. Despite their utility, large animal models are costly and low-throughput, limiting deep mechanistic explorations and functional validation. Alternatively, human primary cardiac slices are useful ex vivo models for studying cardiac physiology¹⁴. State-of-the-art preparation techniques can preserve the human myocardium’s key physiological function and structure while providing expanded throughput¹³. However, primary cardiac samples have limited availability and are incompatible with long-term studies spanning days or weeks.

Human-induced pluripotent stem cells (hiPS cells) may bridge the methodological gap of existing models and provide a scalable platform to test diverse interventions. Specifically, hiPS cells provide a patient-specific and unlimited source of CMs (hiPS cell-derived CMs, hiPS cell-CMs), which can also serve as the building blocks for cardiac organoids and engineered heart tissues (EHTs)^15–18. Cardiac organoids are typically generated through the orchestrated differentiation of human stem cells, with or without the guidance of biomaterials, resulting in multicellular structures with features of native tissue¹⁹. Self-organizing cardiac organoids are particularly useful for studying multicellular interactions in cardiac development and modeling congenital heart disease^20–22. By contrast, EHTs are generated from predifferentiated CMs, nonmyocytes and a hydrogel matrix^18,23–25. Unlike organoids, the resulting three-dimensional (3D) constructs are attached to elastic anchors, mimicking the mechanical load of the human myocardium and facilitating cell alignment. Moreover, the electrical field stimulation of EHTs has gained increasing attention, partly for its potential ability to promote cardiac maturation^26–31. As changes in contractile function are a hallmark of many pathological and physiological adaptations driven by heart rhythm variations, we reason that EHTs are attractive tools for modeling changes in cardiac contractile function and rhythmicity induced by genetic and environmental stress^32–34.

Development of the Protocol

Recently, we published an in vitro workflow to model human heart rhythm in 3D EHTs and recapitulated key features of tachycardia-induced cardiomyopathy previously observable only in large animal models³⁵. This workflow has been systematically optimized over the past several years, from EHT fabrication to the design of the stimulation chamber and pacing parameters.

The EHT fabrication protocol was adapted from a well-established technique that uses silicone racks with elastic pillars^18,36,37. This method was chosen for its reproducibility, consistency and compatibility with our custom pacing system. Of note, to accelerate the maturation of the EHTs and more accurately reflect adult cardiac physiology, we adopted chemical stimulations, including dexamethasone, triiodothyronine and fatty acids^38–40. We confirmed their promaturation effect with functional and molecular characterizations³⁵. We have successfully implemented this end-to-end workflow, generating EHTs across multiple hiPS cell lines from human donors of various sexes, ages and disease backgrounds³⁵.

Based on the dimensions of the EHT racks, we custom-designed and fabricated reusable and autoclavable stimulation chambers with a capacity for 16 EHTs. We also developed a low-cost pacing device that produces programmed patterns of electrical pulses with alternating polarity. Unlike monophasic pacing, biphasic stimulation (stimulation with alternating pulses) avoids electrical toxicity through oxidizing the culture medium. Customized electrical systems capable of generating biphasic stimulation patterns often demand engineering expertise that is less common among biologists, such as the implementation of printed circuit boards (PCBs) and experience with specialized software such as LabView⁴¹. We developed our method to require minimal engineering skills to implement. Specifically, the electrical pacing system can be manually assembled cheaply using easily accessible materials. The combination of an Arduino and an H-bridge circuit offers diverse stimulation patterns with polarity reversibility, enabling the long-term simulation of pathophysiological heart rhythm in EHTs.

Similarly, analyzing EHT contractility frequently requires programming skills in tools such as MATLAB, creating additional barriers to the broader adoption of these technologies outside of the field of biomedical engineering. Therefore, our method also features the analysis of EHT contractility using Tracker, an open-source software, which does not require programming skills. Conceptually, we highlight EHT pacing as a valuable tool for studying heart rhythm-induced cardiac pathophysiology, such as tachycardia-induced cardiomyopathy, which remains an underexplored concept, as most studies focus on cardiac maturation.

Applications of the method

The primary application of this platform is to study how various patterns of beating rhythm affect cardiac pathophysiology at molecular and functional levels in acute (minutes to hours) and chronic (days to weeks) settings. As a proof-of-concept, we demonstrated that sustained tachycardia or an irregular beating rhythm can impair cardiac contractile function in EHTs, consistent with results from animal studies³⁵. Beyond its primary functionality, our Protocol may be adapted for various applications. First, our method may be used to screen for antiarrhythmic drugs. We showed that clinically proven medications for heart rate reduction could blunt the EHT response to pacing-induced contraction³⁵. In addition, this Protocol provides an opportunity to investigate the interaction between genetic stress and environmental stress. For instance, simulated tachycardia in EHTs derived from patients with inherited hypertrophic cardiomyopathy (HCM) resulted in a much greater functional impairment than in EHTs from healthy controls³⁵. Finally, given that heart rhythm alteration is ubiquitous in heart disease, this platform may be adapted to simulate complex cardiac pathophysiology in combination with other biochemical and physical stimulations.

Comparison with other methods

The primary application of our method is to study the impact of the beating rhythm on cardiac pathophysiology. Other existing approaches for such studies can be categorized into three groups.

1. Electrical pacing of primary CMs

Electrical pacing of primary CMs was a common approach before the advancement of hiPS cell technology^5,6,42–46. These methods typically use rodent CMs. However, species differences pose a notable limitation, as resting heart rates in mice and humans differ substantially. Moreover, primary CMs are unsuitable for long-term studies owing to their propensity for apoptosis and dedifferentiation⁴⁷. As a result, it is challenging to model chronic conditions such as tachycardia-induced cardiomyopathy or arrhythmia-induced cardiomyopathy by pacing primary cells.

2. Electrical pacing of 2D hiPS cell-CMs

Advances in hiPS cell technology provide an attractive alternative^48–52. Unlike primary cells, hiPS cell-CMs are patient-specific and adapted for in vitro culture and long-term studies. However, hiPS cell-CMs cultured in two dimensions are generally less mature than those in 3D culture systems⁵³. Furthermore, the lack of structural features resembling native cardiac tissue limits their accuracy in disease modeling. For instance, chronic tachycardia is known to induce the reversible impairment of contractility in patients and dogs and is generally not associated with significant cell death^35,54. Yet, tachypacing in two-dimensional (2D) hiPS cell-CMs induced significant cell apoptosis⁵², a phenotype that is not observed in 3D EHTs or large canine models^35,54.

3. Optogenetic stimulation

In recent years, an increased interest in the optical pacing of hiPS cell-CMs^55,56 and EHTs^57–59 has also been witnessed. Optogenetic stimulation avoids the potential toxicity associated with electrical oxidation and is capable of targeting specific cell populations⁶⁰. Furthermore, compared with 2D culture, EHTs can recapitulate the native cardiac environment, such as cell–cell interactions and dynamic mechanical loading, resulting in more physiologically relevant phenotypes^57,58. However, optical pacing requires the genetic manipulation of the cells and advanced instrumentation, substantially limiting throughput if multiple cell lines are needed.

Limitations

As with any model of human physiology, our method does not capture the full complexity of the biology underlying heart rhythm. First, the sinoatrial node controls normal human heart rhythm, sending electrical impulses via the atrioventricular node to the ventricle and initiating contraction⁶¹. By contrast, our method applies field stimulation that excites the entire EHT. How these different modes of activation affect the downstream biology is unclear. Along the same line, alternating the polarity within or between each pulse is necessary to reduce electrical toxicity. Yet, how closely they mimic the physiological excitation of CMs is unclear. It is worth further optimizing the stimulation waveform in future studies. Second, EHTs or hiPS cell-CMs generally beat spontaneously at ~60 bpm. In most cases, electrical pacing overrides this spontaneous activity, and the contraction of EHT follows the predesigned pattern. However, when the stimulation frequency is substantially lower than the spontaneous rate, it may merely induce extra ‘arrhythmic’ beats on top of the spontaneous contraction. Finally, EHTs lack the cellular and structural complexity of myocardium, including features such as perfusable vasculature. While cardiac tissue with engineered perfusable vasculature has been reported^62–64, these technologies inevitably face other constraints such as scalability, complexity and cost. Despite these limitations, our Protocol provides a novel and valuable tool to dissect the intricate role of heart rhythm on cardiac physiology at functional and molecular levels.

Expertise needed to implement the Protocol

This workflow requires prior skills in hiPS cell culture and cardiac differentiation, which can be implemented by any skilled bench scientists with training in cell culture following established protocols^65,66. While our workflow involves building a pacing device from an Arduino board and circuits, we provided step-by-step instructions to ensure that biologists with minimal or no training in circuitry can manually assemble the setup. Moreover, programming the Arduino board for this Protocol requires minimal skills in coding, as most applications should not require more than 10–20 lines of code, and a variety of ready-to-use programs for users without coding experience are available (Supplementary Code 1).

Experimental design

Our workflow is composed of four modules: derivation of matured EHTs from hiPS cells, preparation of a custom-built pacing platform, electrical stimulation of the EHTs and functional and molecular analysis of the EHTs (Fig. 1). The pacing platform consists of an Arduino microprocessor, which enables versatile stimulation output, an H-bridge circuit to control the polarity of the electrical output and the stimulation chamber to house the EHTs (Fig. 1).

Fig. 1 |. Graphical illustration of modeling heart rhythm using human EHT.

hiPS cell-CMs are used to generate 3D EHTs and matured with triiodothyronine, dexamethasone and fatty acids. A custom-built pacing device, controlled by an Arduino microprocessor, enables the programmed stimulation of the EHTs to mimic various patterns of human heart rhythm. Functional analysis (for example, contraction velocity, relaxation velocity and contraction amplitude) and molecular analysis (for example, transcriptomics and proteomics) allow us to dissect the relationship between human heart rhythm and cardiac pathophysiology. Part of the graphical illustration was adapted from ref. 35, Springer Nature Ltd.

Fabrication and maturation of EHTs (Steps 1–63)

The first part of our Protocol is to generate mature EHTs from hiPS cell-CMs. The generation of hiPS cell-CMs has been described in detail in previous protocols and thus is not included here^65,66. To make EHTs, we mix hiPS cell-CMs, fibrinogen, thrombin and Matrigel solution in a negative mold made from agarose. The mixture undergoes gelation around two elastic pillars. This design enables the later removal of the EHTs from the mold. The cellular and matrix compositions of EHTs are tunable. In this Protocol, we use high-purity hiPS cell-CMs (>85%) as input for EHT fabrication. This can be validated in EHTs after a prolonged culture by immunostaining with the CM marker, troponin T (TNNT2) and nonmyocyte cell marker, vimentin. Vimentin is expressed in fibroblasts and endothelial cells but not in CMs; thus, it can be a useful marker for nonmyocytes in EHTs. We used imaging analysis to confirm that our EHTs contain ~15% vimentin-positive cells, while CMs remain the dominant cell type (Extended Data Fig. 1). Adding supporting cell types, such as endothelial cells and macrophages, into EHTs may better mimic the cellular diversity of the human myocardium and expand the model’s utility^67,68. Yet, these improvements come with disadvantages. Specifically, nonmyocytes complicate the interpretation of results from bulk analyses such as western blotting, proteomics and transcriptomics. Further, nonmyocytes are generally more proliferative than CMs, and it may be challenging to predict the cell composition of an EHT after prolonged culture. To improve the EHT maturity, we treated them with dexamethasone, triiodothyronine and fatty acids for 6 d, which have been reported to promote adult-like phenotypes in electrophysiology, metabolism and contractile function^38,39. This was followed by 3–4 more weeks of culture to mature the tissue further. We reason that maturation is a prerequisite to studying the effect of heart rhythm on cardiac physiology. A fetal heart is better adapted to a fast resting heart rate than an adult heart; thus, the same beating rhythm may have distinct effects on immature and mature EHTs. In alignment with this rationale, a recent study found that tachypacing in early stage EHTs promotes maturation but not in late-stage EHTs²⁸.

Preparation of the pacing apparatus (equipment setup)

Our pacing setup consists of two interconnected parts: a stimulation chamber and a programmable pacing device. The stimulation chamber was custom-designed using Fusion 360 software and manufactured using polycarbonate through computer numerical control machining. It has four separate wells that can house four EHTs in each well (Fig. 2a). Carbon electrodes were inserted into the sides of each well to conduct field stimulation (Fig. 2b). The carbon electrodes should be replaced after a maximum of three uses or immediately following experiments involving drug treatments. The stimulation chamber, cleaned and autoclaved after each use, is reusable for over 20 cycles. The pacing device is an Arduino microprocessor coupled with an H-bridge circuit (Extended Data Fig. 2). This design features several unique advantages. First, our pacing device is portable and cost-effective, costing an average of less than US$50 per unit. It can be manually built in hours without any specialized large instruments. This enables us to build several units and run multiple independent experiments in parallel. Second, our pacing device can generate unlimited stimulation patterns to model diverse heart rhythms. Theoretically, it is possible to specify the interval for every pulse for the entire experiment, whereas commercial pacing instruments typically offer limited preset patterns. Finally, the H-bridge circuit in the device is designed to switch the polarity of stimulation. This technical improvement over prior monophasic stimulation significantly extends the pacing duration to days and weeks, opening up numerous research possibilities⁴¹.

Fig. 2 |. Customized stimulation chamber for EHTs.

a, Computer-aided design of the stimulation chamber for EHTs with the capacity for 16 EHTs. b, A stimulation chamber with carbon electrodes inserted into each well to generate field stimulation. The electrical wires (yellow) are connected to the carbon electrodes. The chamber does not have a lid and needs to be placed in a large Petri dish to maintain sterility. c, Each rack of EHTs is placed in each well of the chamber, with carbon electrodes on both sides of the EHTs. d, Alligator clips connect the stimulation chamber to the pacing device. In this example picture, the top two rows are connected to the circuit and will be stimulated. The bottom two rows remain unpaced.

Electrical pacing of the EHTs (Steps 70–78)

Mature EHTs are transferred from regular culture plates to the stimulation chamber and placed inside the incubator. The stimulation chamber is connected to the pacing device located outside the incubator. The pacing program is written in the Arduino integrated development environment (IDE), a free and easy-to-use coding environment, and uploaded to the Arduino board via a universal serial bus (USB) connection, which will start executing the program when the board is connected to a power source. The Arduino programming language is a C++ code variant. Users can specify the duration of each electrical pulse, the direction of the pulse and the interval after each pulse, which results in unlimited combinations that may resemble different types of heart rhythm. Here, we provide several ready-to-use programs with annotations for users with minimal coding experience. For instance, these programs would allow users to pace EHTs at constant or irregular rates for a designated duration per day (Supplementary Code 1). The contraction of EHTs can be seen through the chamber’s lid, allowing the user to check if the EHTs are capturing the electrical pulses easily. Real-time updates can be made to change stimulation patterns by simply uploading a new program to the Arduino board, which will overwrite the previous program. The stimulation chamber contains enough culture medium for ~2–3 d. If the experimental duration exceeds 3 d, the stimulation chamber can be disconnected from the pacing device and transferred to a biosafety cabinet to change the medium. Doing so allowed us to run pacing experiments lasting from several days to over a month.

Functional (Steps 86–103) and molecular analysis (Steps 104–123) of EHTs

To measure function, EHTs should be transferred into 24-well culture plates, as the stimulation chamber is made of polycarbonate and not optically clear. In our studies, we recorded videos of contractions at 150 frames per second and analyzed the footage using Tracker, a free video analysis and modeling tool. This software allows users to track the displacement of the elastic pillar and generate a data table of locations over time. We then used the data to calculate parameters such as maximum contraction velocity, maximum relaxation velocity, duration of contraction and relaxation, contraction amplitude and beating rate. In addition, knowing the dimensions and Young’s modulus of the elastic pillars enables us to calculate the force generated by the EHTs. Contractility can be repetitively measured at any desired time point after day 35, which allows us to delineate the temporal effect of pacing on cardiac function. EHTs can also be subjected to RNA sequencing (RNA-seq) and proteomic analysis as endpoint measurements for deeper molecular insights. Typically, one EHT is sufficient for either RNA-seq or proteomic analysis, allowing us to obtain both functional and molecular profiles from each EHT sample.

Materials

Biological materials

hiPS cell-CMs

• Three hiPS cell lines (SCVI-273, SCVI-15 and SCVI-591) were used to develop this Protocol. All hiPS cell lines were generated by the Stanford Cardiovascular Institute Biobank and are available upon request. The generation of hiPS cell-CMs from hiPS cells was previously described in detail^65,66. This Protocol uses hiPS cell-CMs from differentiation days 12–20

  ▲ CAUTION Cell lines should be regularly checked for quality (for example, morphological examination and pluripotency marker expression) and contamination (for example, mycoplasma).

  ▲ CAUTION Ethical approval, patient consent and relevant national and institutional regulations apply when generating hiPS cells.

Reagents

EHT fabrication

• TrypLE Express Enzyme (Thermo Fisher Scientific, cat. no. 12605036)

• FBS (Thermo Fisher Scientific, cat. no. 26140079)

• Dulbecco’s modified Eagle’s medium/F12 medium (DMEM/F12; Thermo Fisher Scientific, cat. no. 11320033)

• Rock inhibitor Y-27632 2HCl (Selleck Chemicals, cat. no. S1049)

• Matrigel (Corning, cat. no. 356231)

• Fibrinogen (Sigma-Aldrich, cat. no. F8630)

• Sterile saline solution (Teknova, cat. no. S5812)

• Thrombin (Sigma-Aldrich, cat. no. T7513)

• Agarose (Sigma-Aldrich, cat. no. A9539)

• Endothelial cell growth medium 2 growth medium kit (EGM-2; Lonza, cat. no. CC-3162)

• RPMI 1640 medium (Thermo Fisher Scientific, cat. no. 11875119)

• B27 supplement (50×), serum-free (Thermo Fisher Scientific, cat. no. 17504044)

• Aprotinin (Selleck Chemicals, cat. no. S7377)

EHT culture and maturation

• DMEM (Thermo Fisher Scientific, cat. no. 11885092)

• Penicillin–streptomycin (P–S; Thermo Fisher Scientific, cat. no. 15140122)

• Dimethyl sulfoxide (DMSO; Sigma-Aldrich, cat. no. D2650)

• Dexamethasone (Sigma-Aldrich, cat. no. D4902)

• 3,3′,5-triiodo-ʟ-thyronine sodium salt (T3; Sigma-Aldrich, cat. no. T6397)

• Bovine serum albumin, fatty acid-free (BSA; Sigma-Aldrich, cat. no. A7030)

• Oleic acid-BSA (Sigma-Aldrich, cat. no. O3008)

• Palmitic acid (Sigma-Aldrich, cat. no. P0500)

• Sodium hydroxide solution 10.0 M (NaOH; RICCA Chemical, cat. no. 7470–32)

  ▲ CAUTION NaOH is highly corrosive. Avoid breathing the vapor and skin contact. Handle with care in a chemical fume hood.

• Ultrapure water (Thermo Fisher Scientific, cat. no. 10977015)

Sample preparation for omics analysis

• QIAzol lysis reagent (Qiagen, cat. no. 79306)

  ▲ CAUTION QIAzol is highly toxic. Avoid breathing the vapor and skin contact. Handle with care in a chemical fume hood.

• PBS, pH 7.4 (Fisher Scientific, cat. no. 10010049)

• RNA purification kit (Zymo Research, cat. no. R2062)

• Triethylammonium bicarbonate (Sigma-Aldrich, cat. no T7408)

• Halt protease and phosphatase inhibitor cocktail (Thermo Scientific, cat. no. 78441)

• UltraPure sodium dodecyl sulfate solution, 10% (Thermo Scientific, cat. no. 15553027)

• Lysing matrix D 1.4 mm zirconium-silicate spheres (MPI, cat. no. 1169130-CF)

• Pierce universal nuclease for cell lysis (Thermo Scientific, cat. no. 88700)

Equipment

Plastic consumables

• Pipette tips, 10 μL (CELLTREAT, cat. no. 229015)

• Pipette tips, 20 μL (CELLTREAT, cat. no. 229017)

• Pipette tips, 200 μL (CELLTREAT, cat. no. 229072)

• Pipette tips, 1,000 μL (CELLTREAT, cat. no. 229073)

• 1.5 mL centrifuge tubes (Eppendorf, cat. no. 022431081)

• 15 mL centrifuge tubes (CELLTREAT, cat. no. 667051B)

• 50 mL centrifuge tubes (CELLTREAT, cat. no. 229421)

• T75 culture flasks with filter cap (USA Scientific, cat. no. cc7682–4875)

• Nunc 24-well culture plates (Thermo Fisher Scientific, cat. no. 142475)

  ▲ CRITICAL The dimensions of these 24-well plates are optimal for housing the EHT racks and should not be substituted by similar products.

• Aspirating pipets, 2 mL (CELLTREAT, cat. no. 229262)

• Aspirating pipets, 5 mL (CELLTREAT, cat. no. 229205A)

• Aspirating pipets, 10 mL (CELLTREAT, cat. no. 667210B)

• Aspirating pipets, 25 mL (CELLTREAT, cat. no. 667225B)

• 150-mm cell culture dish (Corning, cat. no. 353025)

• Corning vacuum filter system (Corning, cat. no. 431097)

• Polyethersulfone syringe filters, 0.45 μm pore size (Fisher Scientific, cat. no. 13-100-107)

• BD Lure-Lok syringes without needles (Fisher Scientific, cat. no. 14-823-2A)

• BD general use needles (Fisher Scientific, cat. no. 14-826-52)

• Countess cell counting chamber slides (Thermo Fisher Scientific, cat. no. C10312)

Instruments

• Portable benchtop centrifuge (for example, Eppendorf centrifuge 5425 R, cat. no. 5406000240)

• Centrifuge (for example, Eppendorf centrifuge 5810 R, cat. no. 022625501)

• Autoclave (for example, STERIS AMSCO Steam Sterilizer C Series 16C)

• Analytical balances (for example, Denver Instrument APX-60)

• pH meter (for example, Denver Instrument UB-10 pH meter)

• Hotplate/stirrer (for example, VWR hotplate/stirrer, cat. no. 76549-912)

• Sony SI8000 Cell Imaging System (Sony Biotechnology)

• Biosafety cabinet (for example, Thermo Fisher Scientific 1300 Series A2, cat. no. 1347)

• CO₂ incubator (for example, Thermo Forma Series II CO₂ incubator, cat. no. 3110)

• Water bath (for example, Fisher Scientific water bath, cat. no. FSGPD10)

• Automatic cell counter (for example, Thermo Fisher Scientific Countess 3, cat. no. A49891)

• Fine scissors (Fine Science Tools, cat. no. 14058-11)

• Fine forceps (Fine Science Tools, cat. no. 11412-11)

• Glass beakers, 100 mL (Eisco labs, cat. no. CH200001PK12)

• Glass bottles with lids, 250 mL (Sigma-Aldrich, cat. no. CLS1395250)

• Thermo Scientific NanoDrop (Fisher Scientific, cat. no. ND2000)

• Bead Mill 4 mini homogenizer (Fisherbrand, cat. no. 15-340-164)

EHT and pacing setup

• EHT silicone racks (DiNABIOS, cat. no. C0001)

• Teflon spacers (DiNABIOS, cat. no. C0002)

• Customized stimulation chamber computer numerical control-machined from polycarbonate (Fig. 2a). See the design file of the chamber in Supplementary Data 1

• Momentive RTV108 one-part silicone sealant, translucent (Momentive, cat. no. RTV108–85ML)

• Carbon rods 12 inches × 0.120 inch diameter (Ladd Research, cat. no. 30251)

• Reynolds Wrap aluminum foil (Reynolds Wrap, cat. no. 458742928317)

• Black+Decker Powerconnect cordless electric drill (Black+Decker)

• Tool set with screwdrivers (Cartman)

• Micro twist drill bits, 0.8 mm (Amazon)

• Wire stripper and cutter, 16–26 American Wire Gauge (AWG) (Amazon)

• Breadboard with 830 points (ELEGOO)

• ELEGOO UNO R3 board with USB cable (ELEGOO)

• ELEGOO alternating current/direct current (AC/DC) power adapter for Arduino UNO, 9 V (ELELGOO)

• SoulBay 30 W universal AC/DC adapter switching power supply with eight selectable adapter tips (SoulBay)

• BD139 negative–positive–negative (NPN) transistors (ALLECIN)

• BD140 positive–negative–positive (PNP) transistors (ALLECIN)

• Resistors, 1k Ohm (Amazon)

• Solid electrical wire kit, 22 AWG (Fermerry)

• Solid electrical wire kit, 26 AWG (Fermerry)

• Alligator clips with insulation cover (DROK)

• Multimeter (Klein Tools)

• WAGO 222–413 lever-nuts 2 conductor (WAGO)

• Breadboard and Arduino holder (SunFounder)

• Portable handheld tablet oscilloscope (FNIRSI)

• Optional for researchers without experience in soldering: solderable PCB board, gold-plated (GikFun), and soldering iron kit (Plusivo)

Reagent setup

Matrigel coating solution

Thaw the Matrigel solution at 4 °C overnight. Add 0.4% (v/v) Matrigel to DMEM/F12 medium. Store at 4 °C. Use for ≤2 months.

Fibrinogen stock solution (10×)

Dissolve 250 mg of fibrinogen in 5 mL of saline to yield a 50 mg/mL solution. Heat the mixture in a water bath set at 37 °C to dissolve the powder. Filter the solution using the 0.45 μm polyethersulfone syringe filter. Aliquot the solution into 1.5-mL tubes with 170 μL per tube. This volume will be sufficient for making 16 EHTs. Store at −80 °C and avoid freeze–thaw cycles. Use for ≤6 months.

Thrombin stock solution (33×)

Dissolve thrombin powder in ultrapure water at 100 units/mL in a biosafety cabinet. Aliquot in 1.5-mL tubes with 50 μL per tube. This volume is sufficient for making 16 EHTs. Store at −80 °C and avoid freeze–thaw cycles. Use for ≤6 months.

▲ CRITICAL Thrombin solution should be used within 6 months of reconstitution. Old solutions, even stored at −80 °C, may expire and have significantly reduced activity.

Aprotinin stock solution (2,000×)

Dissolve 33 mg/mL aprotinin powder in DMEM/F12 medium. Aliquot in 1.5-mL tubes with 250 μL per tube. Store at −80 °C and avoid freeze–thaw cycles. Use for ≤1 year.

Replating medium

Prepare 10% (v/v) of FBS and 10 μM of rock inhibitor Y-27632 in DMEM/F12 medium. Filter with a 0.22-μm filter flask. Store at 4 °C. Use for ≤1 month.

RPMI-B27 medium

Add 2% (v/v) B27 into RPMI medium. Store at 4 °C. Use for ≤1 month.

DMEM-B27 medium

Add 2% (v/v) B27 into DMEM medium. Store at 4 °C. Use for ≤1 month.

EGM-2 medium

Thaw the EGM-2 supplements at 4 °C overnight and add all the supplements to the EGM-2 basal medium to obtain the complete EGM-2 culture medium. Store at 4 °C. Use for ≤1 month.

EHT compaction medium

Mix an equal volume of EGM-2 medium and RPMI-B27 medium. Add 0.05% (v/v) aprotinin stock solution and 1% (v/v) P–S. Filter with a 0.22-μm filter flask. Store at 4 °C. Use for ≤1 month.

Palmitate-BSA conjugate (4 mM)

In a 100-mL glass beaker, prepare 12 mM NaOH solution by adding 60 μL of NaOH stock (10 M) into 50 mL of ultrapure water. Add 0.15 g of palmitic acid. Heat the palmitic solution to 80–90 °C on a hot plate. Use a thermometer to monitor the temperature closely. The palmitic acid powder should completely dissolve at ~80–90 °C. Prepare 10% (w/v) BSA solution by adding 10 g of BSA to 100 mL of DMEM/F12 medium. Heat the BSA solution in a water bath to 37 °C. Slowly add 1 volume of palmitate solution to 2 volumes of BSA solution. This should result in a clear solution. Filter with a 0.22-μm filter flask. Store at 4 °C. Use for ≤6 months.

▲ CRITICAL The temperature of the palmitate solution and BSA solution is key to successful conjugation. For instance, if the BSA solution is too cold, palmitate will precipitate instead of conjugating with BSA.

T3 stock solution (1 mM)

Add 6.7 mg of T3 powder to 10 mL of DMSO in a 15-mL tube. Aliquot and store at −80 °C and avoid freeze–thaw cycles. Use for ≤1 year.

Dexamethasone stock solution (24 mM)

Add 94 mg of dexamethasone powder to 10 mL of DMSO in a 15-mL tube. Aliquot and store at −80 °C and avoid freeze–thaw cycles. Use for ≤1 year.

EHT culture medium

Mix an equal volume of DMEM-B27 and RPMI-B27 medium. Add 0.05% (v/v) aprotinin and 1% (v/v) P–S. Filter with a 0.22-μm filter flask. Store at 4 °C. Use for ≤1 month.

EHT maturation medium

In 500 mL of EHT culture medium, add 5 μL of T3 stock, 20.8 μL of dexamethasone stock, 10 mL of palmitate-BSA conjugate and 5 mL of oleic acid-BSA. Filter with a 0.22-μm filter flask. Store at 4 °C. Use for ≤1 month.

Equipment setup

EHT stimulation chamber

● TIMING 2–3 d

1. Cut the carbon rods into pieces of 10.5–11 cm in length.

2. Using an electric drill with a 0.8-mm drill bit, drill a hole ~0.5 cm from one end of each carbon rod. Each rod needs only one hole.

   ▲ CAUTION The 0.8-mm drill bits are brittle and may break during drilling, posing a safety hazard. Protective eyewear should be used. In addition, start with a low drill speed and gradually increase as necessary.

3. Insert eight carbon rods into the four wells of a stimulation chamber. Alternate the orientation of each rod so that the holes on the rods are on the opposite ends in each well (Fig. 2b–d).

4. Seal all eight intersection points between the carbon rods and the chamber using silicone sealant. Leave the chamber in a fume hood for >24 h to allow the sealant to solidify completely.

5. Rinse the chamber with water. Add 10 mL of water to each well of the chamber and place the chamber in a 150-mm Petri dish. Place the dish on a shaker for 1 h. Change the water and repeat the rinsing two more times. If water leakage is observed, apply silicone sealant to the region of leakage, wait for >24 h and restart the rinsing.

6. Wrap the chamber in multiple layers of aluminum foil and autoclave. Use a gravity cycle.

7. In a biosafety cabinet, remove the autoclaved chamber from the aluminum foil.

8. Prepare eight pieces of AWG 22 wires (~10 cm). Strip the two ends using the wire stripper and wipe clean with 70% ethanol.

9. Connect one end of each wire to the carbon rods through the 0.8-mm holes (Fig. 2b–d).

10. Place the chamber in a 150-mm Petri dish and close the lid. The chamber can be stored in a sterile environment (for example, biosafety cabinet) for up to 2 months before use.

Pacing system: H-bridge circuit and Arduino microprocessor

● TIMING 2 h

1. Cut and strip the ends of 26 AWG wires and use them to connect the two positive rails on the 830-point breadboard, and then, connect the two negative rails.

2. Place one BD140 transistor on points h19, h20 and h21. Place another BD140 transistor on points h35, h36 and h37.

3. Place one BD139 transistor on points c19, c20 and c21. Place another BD139 transistor on points c35, c36 and c37.

4. Using 1 k Ohm resistors, make the following connections: g21 to g25, g37 to g41, b21 to b25 and b37 to b41.

5. Cut and strip the ends of 26 AWG wires and make the following connections: connect i19 to the positive rail, i35 to the positive rail, b19 to the negative rail, b35 to the negative rail, f25 to e25 and f41 to e41. The completed H-bridge circuit is shown in Extended Data Fig. 2a.

6. Cut two pieces of 22 AWG wires (~1 m long) and strip the ends. Connect one end of each wire to the H bridge circuit at points a20 and a36, respectively. Connect the other end of each wire to an alligator clip (Extended Data Fig. 2b). These two wires will provide the electrical output for pacing EHTs and will be connected to the stimulation chamber.

7. Connect a universal power adapter to a female terminal connector (green color) provided with the adapter.

8. Cut and strip the ends of 26 AWG wires and use them to connect the terminal connector to the positive and negative rails on the breadboard (Extended Data Fig. 2c). This will provide power for the pacing device. The voltage set on the power adapter will determine the voltage applied to the EHTs. Set the power adapter connected to the H-bridge circuit to 5–9 V.

9. Connect the Arduino to the H-bridge circuit. Use 26 AWG wires to connect pin 12, pin 9 and ground (GND) pin on the Arduino to points a25, a41, and the negative rail on the breadboard, respectively (Extended Data Fig. 2d).

10. Connect the Arduino board to the ELEGOO AC/DC power adapter (white color) for Arduino (Extended Data Fig. 2e). This will be the power source for the Arduino.

11. Install the pacing device onto an Arduino/breadboard holder using screws and adhesives (Extended Data Fig. 2f).

12. The complete pacing platform, including the stimulation chamber, the Arduino board and the H-bridge circuit assembled on a breadboard, is shown in Fig. 3a–d. When the system is in active use, place the stimulation chamber (Fig. 3b) in the incubator, and leave the rest of the system outside. Output wires, shown in Fig. 3a, are thin enough to go through the crevice between the incubator door hinges without breaking its seal.

    ▲ CRITICAL For researchers with experience in soldering, the circuit on the breadboard can be soldered onto a solderable PCB board. This will reduce the setup size and further secure all connection points. The circuit diagram for the pacing system, which indicates the direction of the electrical current under different conditions, is shown in Extended Data Fig. 3a–c.

Fig. 3 |. Overview of the setup.

a, The system comprises three interconnected parts: an H-bridge circuit manually assembled on a breadboard, an Arduino microprocessor and the stimulation chamber. b, A stimulation chamber houses up to 16 EHTs from four racks. The chamber is connected to the H-bridge circuit via alligator clips. c, The Arduino board is connected to the H-bridge circuit via three points: pin 9, pin 12 and the ground pin. d, H-bridge circuit on a breadboard enables polarity switch. Signals from pin 9 and pin 12 dictate the direction of current flow. Optionally, the entire circuit could be soldered on a solderable PCB manually.

Procedure

Day −3: replating of hiPS cell-CMs

● TIMING 3 h

• 1Prepare three six-well plates of purified and confluent hiPS cell-CMs between differentiation days 12–20 (refs. 65,66). If the cell confluency in each plate is low, then more plates may be required.
• 2Prepare 100 mL of Matrigel coating solution and 100 mL of replating medium.
• 3Prepare three T75 flasks by adding 8 mL of Matrigel coating solution into each flask and incubating at 37 °C for >1 h. Alternatively, this incubation step can be done overnight at 37 °C.
• 4Remove the old culture medium from hiPS cell-CMs and add 1 mL of TrypLE Express enzyme to each well.
• 5Incubate at 37 °C for 10–20 min.

  ▲ CRITICAL STEP Check the cells under a microscope to ensure they have shrunk into spheres before proceeding to the next step. Forcefully aspirating the cells off the plate increases cell death.

• 6The cells will be in suspension; carefully flush and collect the cell suspension off the plate.

  ◆ TROUBLESHOOTING

• 7In a 50-mL conical tube, mix the cell suspension (~18 mL) with 20 mL of replating medium.
• 8Centrifuge to spin down the cells at 200g for 5 min.
• 9Remove the supernatant and resuspend the cells in 30 mL of the replating medium.
• 10Remove the coating solution from the T75 flasks.
• 11Add 10 mL of cell suspension into each flask and place them in a CO₂ incubator.

Day −2: medium change for hiPS cell-CMs

● TIMING 30 min

• 12Prepare 30 mL of RPMI-B27 medium.
• 13Remove the old medium from the T75 flasks.
• 14Add 10 mL of RPMI-B27 medium to each flask and place them back into the incubator.

Day 0: generation of EHTs from hiPS cell-CMs

● TIMING 4–5 h

• 15Check the confluency of the cells under a microscope. The cells should be >95% confluent.

  ◆ TROUBLESHOOTING

• 16Thaw 1 aliquot of thrombin and 1 bottle of Matrigel on ice. Alternatively, Matrigel may be thawed overnight at 4 °C.
• 17Prepare 100 mL of replating medium.
• 18Prepare 16 sterile 1.5 mL tubes in a biosafety cabinet by dispensing 3 μL of thrombin into each tube. Leave the tubes on ice for later use.
• 19Prepare 100 mL of a 2% (w/v) agarose solution in ultrapure water in a 250 mL glass bottle.
• 20Place 4 EHT silicone racks and 4 Teflon spacers in a clean, empty tip box and wrap with foil.
• 21Autoclave the agarose solution, the EHT racks and the Teflon spacers. Use the liquids cycle.
• 22Transfer the autoclaved supplies into a biosafety cabinet.

  ▲ CRITICAL STEP Steps 22–24 are time sensitive. The autoclaved agarose solution should be used as soon as possible. The solution’s viscosity increases as it cools and will be difficult to pipette.

• 23In a 24-well plate, add 1.6 mL of the hot agarose solution per well to the plate. Use the middle four columns and leave the eight wells on the two sides empty.
• 24Insert the Teflon spacers into the agarose solution (Fig. 4a). Set the plate aside for later use.
• 25Transfer the three T75 flasks with hiPS cell-CMs into a biosafety cabinet, remove the old culture medium, and add 10 mL of Tryple Express enzyme into each flask.
• 26Incubate at 37 °C for 10–20 min.
• 27Pipette and flush the cells off the flasks.
• 28Mix the cell suspension (~30 mL) with 20 mL of replating medium in a 50 mL conical tube.
• 29Centrifuge to spin down the cells at 200g for 5 min.
• 30Remove the supernatant and resuspend the cells in 10 mL of replating medium.
• 31Prepare for cell counting. Mix 10 μL of trypan blue and 10 μL of the cell suspension.
• 32Load 10 μL of the mixture onto a Countess cell counting chamber slide.
• 33Load the slide into the Countess 3 automatic cell counter to calculate cell concentration.
• 34On the basis of the cell count, transfer 3.4 ×10⁷ hiPS cell-CMs into a new 50-mL tube. This will be sufficient for 16 EHTs, each using ~2 ×10⁶ cells.
• 35Centrifuge to spin down the cells at 200g for 5 min.
• 36Resuspend the cells in 1.36 mL of replating medium. Keep the cells on ice for the rest of the EHT fabrication process.
• 37Thaw one aliquot of fibrinogen solution at 37 °C. This should take ~3–5 min.
• 38Add 170 μL of cold Matrigel to cell suspension and pipette to mix.
• 39Add 170 μL of fibrinogen solution and pipette to mix. The total volume is now ~1.7 mL. Leave the cell suspension on ice.
• 40Gently remove the Teflon spacers from the 24-well plate (prepared in Step 24) (Fig. 4b).

  ◆ TROUBLESHOOTING

• 41Insert the autoclaved EHT racks into the slots created by the Teflon spacers (Fig. 4c).
• 42Add 100 μL of the cell–gel mixture to a tube with a thrombin aliquot (prepared in Step 18).

  ▲ CRITICAL STEP During Steps 42–45, avoid moving the EHT silicone racks, as this will result in uneven gelation.

• 43Mix quickly and transfer the mixture into the negative mold in the 24-well plate (Fig. 4d).

  ▲ CRITICAL STEP Avoid generating air bubbles during pipetting. Bubbles can be trapped in EHTs, which affects their functional and mechanical properties.

• 44Repeat Step 43 until all 16 slots are filled with the cell–gel mixture.
• 45Gently transfer the plate into a CO₂ incubator. Incubate at 37 °C for 1.5 h.
• 46Warm up 10 mL of replating medium at 37 °C.
• 47Add 0.5 mL of the warm replating medium into each well that contains the EHT.
• 48Incubate at 37 °C for >15 min.
• 49Prepare 100 mL of EHT compaction medium.
• 50In a new 24-well plate, add 1 mL of EHT compaction medium in each well of the middle four columns. This should match the layout of the EHT plate.
• 51Gently remove the EHT racks (Fig. 4e) from the 24-well plate and transfer them to the new plate containing the EHT compaction medium.

  ◆ TROUBLESHOOTING

• 52Return the plate to the CO₂ incubator.

Fig. 4 |. Steps of EHT fabrication.

a, The insertion of Teflon spacers into a 24-well plate containing hot agarose solution. b, The removal of the Teflon spacers creates a negative mold for EHTs. c, Insertion of sterile EHT silicone racks. d, The injection of the cell–gel mixture containing hiPS cell-CMs, fibrinogen, Matrigel and thrombin. e, The removal of the EHTs from the mold after complete gelation, which takes ~1.5–2 h. f, A representative image of a well-formed EHT after >2 weeks of culture showing successful compaction as indicated by a narrow body and round ends. Scale bar, 1 mm.

Day 1–15: EHT compaction

● TIMING medium change every 2 d: 15 min

• 53Transfer the EHTs to the empty wells on the two sides of the plate in a biosafety cabinet.
• 54Remove the old medium.
• 55Add 1 mL fresh EHT compaction medium.
• 56Transfer the EHTs back to the wells containing fresh medium.
• 57Return the plate to the incubator and repeat media changes (Steps 53–57) every 2 d. In 2 weeks, EHTs should compact (Fig. 4f) and generate synchronized contractions.

  ◆ TROUBLESHOOTING

Day 15–21: EHT maturation

● TIMING medium change every 3 d: 15 min

• 58Prepare 35 mL of maturation medium.
• 59Transfer the EHTs to the empty wells on the two sides of the plate in a biosafety cabinet.
• 60Remove the old EHT compaction medium.
• 61Add 1 mL fresh EHT maturation medium.
• 62Transfer the EHTs back to the wells containing fresh medium.
• 63Return the plate to the incubator and repeat media changes (Steps 58–63) every 3 d.

Day 21–35: EHT maintenance

● TIMING medium change every 3 d: 15 min

• 64Prepare 250 mL of EHT culture medium.
• 65In a biosafety cabinet, transfer the EHTs to the empty wells on the two sides of the plate.
• 66Remove the old medium.
• 67Add 1 mL of fresh EHT culture medium.
• 68Transfer the EHTs back to the wells containing the fresh medium.
• 69Return the plate to the incubator and repeat media changes (Steps 64–69) every 3 d. Examine EHT morphology and contraction daily using a benchtop microscope.

  ◆ TROUBLESHOOTING

Day 35: setting up programmed pacing of EHTs to emulate heart rhythm

● TIMING 1 h

• 70Prepare 40 mL of EHT culture medium and warm up in a water bath at 37 °C.
• 71Add 10 mL of EHT culture medium into each well of the EHT stimulation chamber (see ‘Equipment setup’ for more details).
• 72Transfer the EHTs into the stimulation chamber and close the lid. Place the chamber in an incubator to allow the EHTs to recover.
• 73Download the free Arduino IDE onto a laptop.
• 74Connect the Arduino board to the laptop using a USB type A to B cable. This will allow users to upload the pacing program to the Arduino.
• 75For users without experience in coding, copy and paste the desired stimulation pattern into the IDE (Supplementary Code 1). Upload the code to the board. The board will hold one program at a time. Uploading a new program will rewrite the previous program. Users with experience in coding may choose to write their stimulation programs.
• 76After uploading the program, unplug the USB from the Arduino board.
• 77Connect the pacing device to the stimulation chamber through the alligator clip (Fig. 3a,b). The pacing device should be placed outside the incubator near a power plug, and the wires can go through the space between the door hinges.

  ▲ CRITICAL STEP Do not place the pacing device inside the incubator, as the humidity and high temperature may damage the electronics.

• 78
  Plug in the power supplies for both the H-bridge circuit and Arduino. This will start pacing the EHTs. The contraction of the EHTs can be visually checked by the naked eye to confirm their response to pacing.

  • For long-term pacing experiments, culture medium needs to be changed at least every 3 d, as instructed in Steps 79–85. EHT contractility measurement and analysis (Steps 86–103) can be performed at any desired time point after day 35. For endpoint molecular profiling, proceed to sample preparation for transcriptomic analysis (Steps 104–114) and/or proteomic analysis (Steps 115–123).
    ◆ TROUBLESHOOTING

Day 37: medium change between pacing

● TIMING medium change every 3 d: 30 min

• 79Unplug the power supply for both the Arduino and the H-bridge circuit.
• 80Disconnect the alligator clips from the EHT stimulation chamber.
• 81Transfer the chamber into a biosafety cabinet.

  ◆ TROUBLESHOOTING

• 82Transfer the EHTs into a new 24-well plate temporarily.
• 83Remove the old medium from the stimulation chamber and add 10 mL of fresh EHT culture medium to each well.
• 84Place the EHT racks back into the chamber and return the chamber to the incubator. Allow the EHTs to recover for 10–15 min.
• 85Reconnect the alligator clips to the stimulation chamber and plug in the power to resume the pacing. Repeat media changes (Steps 79–85) every 3 d.

Recording of EHT contractility

● TIMING 1.5 h

• 86Warm up 20 mL of EHT culture medium to 37 °C.
• 87Prepare a new 24-well plate.
• 88Add 1 mL of warm medium into each well.
• 89Transfer the EHTs from the stimulation chamber into the 24-well plate.
• 90Turn on the microscope, camera and incubation system 20 min before use. Set the heating unit temperature to 37 °C.
• 91Place the 24-well plate with the EHTs on the plate holder and allow them to stabilize for 15 min before recording.
• 92In the SONY SI8000 software, set the camera frame rate to 150 frames per second and the video/image format to 1,024 × 1,024 pixels. The total recording time is 10 s. Use the 10× objective.

  ▲ CRITICAL STEP Alternatively, recording of EHT contractility (Steps 90–95) could be implemented using any microscope platform with a high-speed video camera and an incubation system.

• 93Select regions of interest (ROIs) for recording. For each EHT, select one ROI near each elastic post. Displacement of the elastic posts will be used for later analysis.
• 94Set the saving directory to an external drive with >100 GB of storage.
• 95Start recording and avoid disturbing the imaging station during recording.

  ■ PAUSE POINT Recorded videos from different experimental groups and batches can be stored in a large drive and later analyzed together (Steps 96–103). This helps reduce variation.

Analysis of EHT contractility

● TIMING 1.5 h

• 96Install Tracker, a free video motion analysis tool, on a laptop.

  ▲ CRITICAL STEP Tracker offers both MacOS and Windows versions, and the following instructions were based on the MacOS version. Minor changes might be needed if using Windows.

• 97Import the EHT contraction videos obtained from Steps 86–95 into Tracker.
• 98(Optional) If functional data needs to be presented as absolute values, create a calibration stick in the software and specify the actual dimensions of the video frame.
• 99Create a new point mass under the ‘Track’ menu.
• 100Hold down the ‘command’ and ‘shift’ keys and click on the feature in the video to be tracked. The Autotracker window will pop up, and an ROI will be generated.
• 101Click ‘Search’. The video will start playing, and the software will track the ROI’s trajectory.

  ◆ TROUBLESHOOTING

• 102Upon completion of tracking, open the ‘Plot’ window to retrieve parameters such as contraction amplitude, maximum contraction velocity and maximum relaxation velocity. The beating rate can also be calculated by reading the interval between peaks.
• 103(Optional) To perform a more detailed analysis, export raw trajectory data. This facilitates the calculation of more customized parameters using tools such as MATLAB or Python.

Sample preparation for transcriptomic analysis

● TIMING 1.5 h

• 104Gently remove the EHTs from the silicone racks using surgical tweezers.
• 105Transfer each EHT into a clean 1.5-mL tube.
• 106Add 1 mL of PBS into each tube to rinse the EHT and remove the PBS.
• 107Add 0.5 mL of PBS.
• 108Use a fine pair of scissors to cut the EHT into small pieces.
• 109Centrifuge to spin down the tissue chunks at 200g for 5 min.
• 110Gently remove the supernatant.
• 111Add 400 μL of QIAzol lysis reagent into each tube and pipette to lyse the tissue.

  ▲ CAUTION QIAzol is toxic if inhaled or swallowed and should only be handled in a chemical fume hood. All waste, such as tips contaminated by QIAzol, needs to be disposed of in designated containers.

  ■ PAUSE POINT The resulting tissue lysate may be frozen at −80 °C for long-term storage.

• 112Follow the manufacturer’s instructions for RNA extraction using a Zymo RNA purification kit.
• 113Measure the RNA concentration using a NanoDrop.
• 114Store the extracted RNA at −80 °C for long-term storage or proceed to RNA sequencing.

Sample preparation for proteomic analysis

● TIMING 1.5 h

• 115Gently replace EHT media with 1 mL of PBS.
• 116Remove PBS and replace with new PBS. Repeat twice for a total of three washes.
• 117Gently remove the EHTs from the silicone racks using surgical tweezers.
• 118Combine two EHTs into one clean 1.5 mL tube and place the tube on ice.
• 119Add 500 μL of lysis buffer (10% sodium dodecyl sulfate, 100 mM triethyl ammonium bicarbonate, 1× protease and phosphatase inhibitor and Pierce universal nuclease diluted 1:5,000).
• 120Add zirconium-silicate spheres equal to ~1/2 of the total 500 μL volume.
• 121Use a mini bullet blender to lyse the EHTs (alternate 30 s at 5 m/s speed and 30 s on ice until large tissue chunks are no longer observed and the mixture appears homogeneous).
• 122Centrifuge for 10 min at 12,000 rpm at 4 °C to pellet debris and remove bubbles.
• 123Carefully pipette the lysate into a new 1.5 mL tube and store it at −80 °C until peptide preparation.

Troubleshooting

Troubleshooting advice can be found in Table 1.

Table 1 |.

Troubleshooting table

Step	Problem	Possible reasons	Solution
6	hiPS cell-CMs are difficult to flush off the plate	High cell density and excessive extracellular matrix	Extend the digestion time to over 20 min; alternatively, use TrypLE Ex 10×, a concentrated version of TrypLE Ex
15	hiPS cell-CMs are sub-confluent (<80% confluency) 3 d after replating	Excessive cell death during replating at Steps 1–11	Prepare fresh replating medium and optimize digestion time at Step 5. Ideally, cells should be easy to flush off and disperse into single cells suspension after digestion. Underdigested cells require more vigorous pipetting, which may cause reduced cell viability
40	Agarose molds break when Teflon spacers are removed	Incomplete gelation of agarose	Extend the cooling time before removing the spacers
51	Tissue breaks off the silicone post during transfer	Expired or inactive thrombin	Prepare fresh thrombin aliquot from powder
57	EHTs fail to compact and do not generate synchronized contraction	Insufficient hiPS cell-CMs	Increase the quantity of input hiPS cell-CM for EHT fabrication
		Excessive cell death during EHT fabrication	Use freshly made replating medium and optimize the digestion time during EHT fabrication
	EHTs compact successfully but do not generate synchronized contraction	Overgrowth of nonmyocytes such as fibroblasts in the tissue, resulting in high internal stress and tension	Increase the purity of input hiPS cell-CMs via metabolic purification (for example, glucose starvation)
69	EHTs become very thin and break from the middle	Insufficient amount of aprotinin leading to rapid degradation of fibrin and Matrigel	Increase the dose of aprotinin in the medium by 100–300%
	EHTs break off the silicone post during long-term maintenance	Overgrowth of nonmyocytes such as fibroblasts in the tissue, resulting in high internal stress and tension	Increase the purity of input hiPS cell-CMs via metabolic purification (for example, glucose starvation)
		EHT hypercontractility	Use compaction medium for maintenance because it contains reduced calcium than EHT culture medium, which would lower contractile force
78	EHTs do not respond to the designed pacing program	Insufficient electrical current to activate contraction	Increase input voltage by adjusting the setting on the power adapter that is connected to the H-bridge circuit
			Increase the duration of each pulse to up to 10 ms
		Slow contraction/relaxation kinetics of the EHTs prevents them from beating at a high frequency	Gradually increase the pacing frequency. For instance, if a 3 Hz beating rate is desired, pace at 1.5 Hz for 30 min, then increase the frequency by 0.5 Hz every 30 min until the desired beating rate is reached
		Unknown reasons	Before the start of pacing, incubate the EHTs in the stimulation chamber for an extended time (>1 h) to allow them to adapt and recover from the transfer
81	Culture medium turns excessively acidic and yellow during pacing, resulting in unexpected cell death in EHTs	Asymmetrical electrical output, resulting in electrolysis of the culture medium	Use an oscilloscope to check for the waveform of the electrical output. The signal should be alternating evenly (Fig. 5). If not, check the wire connections in the pacing device for misconnections or loose wires and check the Arduino code for bugs
101	Autotracker fails to follow the motion of the ROI	Video is unfocused and blurry	Rerecord the video to ensure that EHTs are in focus with moderate lighting. Avoid overexposure as it bleaches texture features necessary for ROI tracking
		Pixels in the selected ROI do not have sufficient features to be tracked between frames	Select new ROIs
			Optimize the ‘Evolve’ rate in the tracking window

Timing

Equipment setup, EHT stimulation chamber: 2–3 d

Equipment setup, pacing device setup: 2 h

Steps 1–11, replating of hiPS cell-CMs: 3 h

Steps 12–14, medium change for hiPS cell-CMs: 30 min

Steps 15–52, generation of EHTs from hiPS cell-CMs: 4–5 h

Steps 53–57, EHT compaction: medium change every 2 d: 15 min

Steps 58–63, EHT maturation: medium change every 3 d: 15 min

Steps 64–69, EHT maintenance: medium change every 3 d: 15 min

Steps 70–78, setting up programmed pacing of EHTs to emulate heart rhythm: 1 h

Steps 79–85, medium change between pacing every 3 d: 30 min

Steps 86–95, recording of EHT contractility: 1.5 h

Steps 96–103, analysis of EHT contractility: 1.5 h

Steps 104–114, sample preparation for transcriptomic analysis: 1.5 h

Steps 115–123, sample preparation for proteomic analysis: 1.5 h

Anticipated results

This Protocol describes the generation of 16 EHTs, which can be scaled up or down depending on experimental needs. Immediately following EHT fabrication (Step 51), EHTs should appear opaque (Fig. 4e), which is an indicator of successful gelation. After 24 h, microscopic contractions should be observed inside the EHTs, signaling good cell viability and recovery. In 1–2 weeks, EHTs should continue to compact and form a ‘dumbbell-shaped’ morphology (Fig. 4f) and produce synchronized contractions capable of displacing the elastic posts (Supplementary Videos 1, 2, 3 and 4). After the maturation treatment (Steps 58–63), EHTs should have increased maturity at the molecular and functional levels. Specifically, our qPCR analysis confirmed that the maturation treatment increased the expression of various maturity markers, such as FABP3, CKMT2 and TNNI3 (ref. 35). In addition, using the method described in Steps 86–103, we confirmed that maturation treatment increased EHT contractile function³⁵. Moreover, EHTs are expected to exhibit pharmacological responses to common cardiac drugs such as calcium channel blockers and beta-blockers. Lastly, if hiPS cell-CMs from disease cell lines are used, the resulting EHTs are likely to present patient-specific phenotypes, which have been reported by others⁶⁹ and us³⁵. In our previous study, we simulated chronic tachycardia in EHTs derived from healthy donors and donors with inherited HCM. Interestingly, over 50% of HCM EHTs completely stopped contractions after prolonged tachypacing, a unique phenotype not observed in wild-type EHTs³⁵. This result mirrors the clinical observation that patients with HCM tend to have an increased risk of sudden cardiac death⁷⁰.

Our Protocol provides detailed instructions to manually build a cost-effective, portable and programmable pacing device. The combination of Arduino and the H-bridge circuit enables us to generate unlimited biphasic stimulation patterns. We used an oscilloscope to visualize the pattern of the electrical output (Fig. 5a–d). Specifically, we could modulate the stimulation frequency by modifying the intervals between pulses (Fig. 5a), modulate the amplitude by adjusting the voltage setting on the power adapter (Fig. 5b), modulate the width of each pulse by changing their duration (Fig. 5c) and modulate the regularity of stimulation pattern by varying the intervals (Fig. 5d). We recommend checking the stimulation pattern using an oscilloscope before pacing the EHTs to ensure it matches the desired program as specified by the code. Alternatively, users may connect the alligator clips (Fig. 3b) to light-emitting diodes to quickly check the direction and frequency of the electrical output, as demonstrated in Supplementary Video 5. While the assembly of the pacing system does not require engineering expertise, we provide the circuit diagram for those who wish to further improve or customize the system (Extended Data Fig. 3).

Fig. 5 |. Representative stimulation patterns from the customized pacing device.

a–d, The Arduino program enables modulation of pacing frequency (a), modulation of the input voltage (b), modulation of the pulse width (c) and irregular pacing to mimic various types of arrhythmia (d). An oscilloscope measured the electrical output of the pacing device. The signal indicates voltage over time. Stimulation polarity was switched after each pulse.

Mature EHTs with a high degree of cell alignment (Fig. 6a) are expected to beat according to the programmed rhythm in response to electrical pacing. Typically, pacing overrides the spontaneous rhythm of the EHTs. Our findings indicate that EHTs can reliably follow pacing frequencies between 1 and 3 Hz (Fig. 6b,c). However, their response becomes less consistent at frequencies above 3 Hz. We found that contractility improves as the pacing frequency increases from 1 to 3 Hz, and significantly decreases at 4 Hz (Fig. 6b,c). Interestingly, 3 Hz (or 180 bpm) is close to the maximum heart rate of adult humans⁷¹, underscoring the physiological relevance of this model. In addition to pacing at different rates, we have been able to model more complex cardiac rhythms in the system. Using the SONY SI8000 analyzer, we generated velocity-time plots from videos (Fig. 7a–d). In these plots, each cycle of the ‘heartbeat’ is indicated by two adjacent peaks: a contraction peak followed by the relaxation peak. Normally, the contraction peak should have a greater amplitude than the relaxation peak. At baseline without pacing, EHT contraction mimics a normal rhythm (Fig. 7a). With programmed pacing, we have been able to mimic sustained tachycardia (Fig. 7b), transient tachycardia (Fig. 7c) and bigeminy (Fig. 7d). A set of ready-to-use stimulation programs is also provided in this Protocol with detailed annotations for users with minimal coding expertise (Supplementary Code 1).

Fig. 6 |. Mature EHTs are highly responsive to electrical pacing.

a, Immunostaining of cryosectioned EHT samples revealed increased cell alignment, a feature reminiscent of the native myocardium. TNNT2 is shown in green, and the cell nucleus (DAPI) is shown in blue. Scale bars, 200 μm (left) and 10 μm (right). b, Measurements of beating rate, maximum contraction velocity and maximum relaxation velocity of an EHT in response to various pacing frequencies. The EHT responded to stimulation at up to 4 Hz but exhibited reduced contractility at this frequency. c, Representative motion traces of the EHT paced at different frequencies.

Fig. 7 |. Simulation of arrhythmias in EHTs.

a, The beating rhythm of unpaced EHTs resembles the normal human rhythm. b–d, EHTs were subjected to programmed electrical stimulation to induce varying beating patterns that resemble chronic tachycardia (b), transient tachycardia (c) and ventricular bigeminy (d). The data are visualized using the SONY SI8000 Analyzer, showing motion velocity over time.

This workflow is also compatible with common omics analysis, for which sample preparations are provided. Each EHT is expected to provide >500 μg of total RNA, sufficient for RNA-seq analysis. For proteomic analysis, EHTs provide >1,000 ng of material even after suspension-trapping peptide preparation, which is well-suited for common data-dependent acquisition methods. We expect to find differentially expressed genes and proteins induced by different pacing programs in the EHTs (Fig. 8a,b). For example, if sustained tachypacing is applied, the EHTs are expected to become hypoxic and transcriptionally upregulate the hypoxia signaling pathway and glycolysis pathway³⁵. An EHT could first undergo functional analysis (Steps 86–103), followed by transcriptomic or proteomic analysis, thereby providing the opportunity to establish a link between molecular changes and functional changes for each sample.

Fig. 8 |. Proteomic and transcriptomic changes induced by simulated tachycardia.

a, A volcano plot of differentially expressed proteins induced by 5 d of sustained tachypacing at 3 Hz. Tachypacing reduced the expression of sarcomeric proteins such as MYH6 and MYBPC3. n = 4 EHTs per group. b, Volcano plot of differentially expressed genes induced by 3 d of sustained tachypacing at 3 Hz. Tachypacing upregulated the expression of hypoxia genes such as MB and LDHA. For each group, n = 2 independent batches that were pooled from 4 EHTs.

Extended Data

Extended Data Fig. 1 |. Confocal images of EHT.

Cryo-sections of EHTs were immunostained for cardiomyocyte marker (TNNT2, shown in green), nonmyocyte marker (vimentin, shown in red) and nuclei (DAPI, shown in blue). Scale bar, 50 μm.

Extended Data Fig. 2 |. Assembly of the pacing device.

a, Custom-built H-bridge circuit on a breadboard with 830 points. b, The connection of the output wires to the H-bridge circuit. The alligator clips, shown in red and black, will be connected to the stimulation chamber. c, The power source (black adapter) connection to the H-bridge circuit. This power source determines the voltage of electrical stimulation. d, The connection of the Arduino microprocessor to the circuit. Pacing programs will be uploaded to and stored in the Arduino. Arduino requires an independent power source (power adapter in white). e, An overview of the completed pacing device. f, The Arduino and the breadboard are installed on a specialized holder (shown in blue).

Extended Data Fig. 3 |. Circuit diagram of the pacing system.

a, When both pins 9 and 12 are off, the circuit is open and there is no current flowing through the load (EHTs). b, When pin 9 is on, and pin 12 is off, the current passes the EHTs from left to right (shown in red). c, When pin 9 is off and pin 12 is on, current passes the EHTs from right to left (shown in blue).

Supplementary Material

The online version contains supplementary material available at https://doi.org/10.1038/s41596-025-01217-w.

Key points.

• The programmable stimulation of mature engineered heart tissue (EHT) emulates diverse patterns of heartbeats associated with physiological and pathological changes, providing a platform to dissect the intricate role of heart rhythm on cardiac physiology at functional and molecular levels.

• This procedure covers EHT fabrication and maturation; the construction of a cost-effective, portable and programmable pacing system; ready-to-use example code; the functional analysis of EHTs and sample preparation for proteomic and transcriptomic analysis.

Data availability

All source data and design files are provided within the paper. RNA-seq data are available at the National Center for Biotechnology Information Gene Expression Omnibus repository, under accession number GSE242727. Source data are provided with this paper.

Code availability

Custom code is provided within the paper.